Home / Publications / News and analysis / ForIntel Research / Market Research With a Source for Every Number

Market Research With a Source for Every Number

Why the market research you can act on is the kind that shows its sources, how to tell sourced research from confident guessing, and what to ask for before you pay for a report.

By Foragentis teamPublished 2026-11-143 min read

Short answer: the market research you can actually act on is the kind that names a source for every number. Sourcing is what lets you check a finding, judge how current it is, and defend it when someone pushes back. Research that hands you confident conclusions with no sources is just an opinion, and you only learn it was wrong after you have bet on it.

Most market research fails not because the numbers are wrong but because you cannot tell whether they are right. A slide deck full of confident figures and no citations is impossible to check. You either trust it whole or throw it out, and both are bad ways to make a decision. The fix is a discipline, not a secret: a source for every number.

Why sourcing is the whole game

A number without a source is a number you cannot defend. When a finding names where it came from, three things become possible:

  • You can check it. Follow the source and confirm the figure means what the report says it does.
  • You can judge its age. A number from last quarter and a number from three years ago are not the same number, and the source tells you which you have.
  • You can stand behind it. When your partner, your lender, or your board asks "where did this come from," you have an answer.

What to do with this: on the next report you are handed, pick the single most important number and try to trace it. If you cannot, you do not have research. You have a claim.

Not all evidence is equal, and good research says so

A trustworthy report does more than cite sources. It tells you how strong each finding is. A figure drawn from a large, direct dataset deserves more weight than one inferred from a handful of mentions, and labeling that difference is what lets you act correctly: lean on the strong findings, treat the thin ones as leads to check rather than facts to bet on.

Why a human still reads it

Collecting data at scale is fast and blind to context. A figure can be technically correct and still misleading, a source can look authoritative and not mean what it seems to. A person reading the findings before they reach you catches those, and flags the gap worth knowing about. Scale plus a human read is what turns raw data into something you can use.

How ForIntel fits

ForIntel is built on exactly this standard. Its reports collect data at scale, each is read by a person before it is delivered, every finding names its sources, and each is labeled by the strength of its evidence. You get research you can trace, weigh, and defend, not a confident guess. See the ForIntel report catalog for the reports available and the sources each one draws on.

FAQ

Why does a source for every number matter in market research?

Because a number you cannot check is a number you cannot defend. When a finding names its source, you can see where it came from, judge how current it is, and stand behind it when someone asks. A report without sources asks you to trust it on faith, and you only find out it was wrong after you have acted on it.

What does it mean to label a finding by strength of evidence?

It means the report tells you how sure to be. A figure drawn from a large, direct dataset is stronger than one inferred from a small sample or a single mention, and a good report says which is which. That lets you lean hard on the solid findings and treat the thin ones as worth a closer look rather than a decision.

Should a person review machine-collected research before I get it?

Yes. Data collected at scale is fast but blind to context. A person reading the findings before delivery catches the figure that is technically correct but misleading, the source that does not mean what it appears to, and the gap worth flagging. Collection at scale plus a human read is what turns raw data into research you can use.